Evaluating the Impact of Ambient AI on Documentation Efficiency and Clinician Burnout in Primary Care Settings

This pilot study aims to evaluate the effectiveness of an ambient listening AI product in improving clinical documentation efficiency and patient satisfaction, reducing clinician burnout in primary care settings, and improving operational implementation strategies by harnessing end-user psychology. Employing a randomized, prospective design, the study involved 25 clinicians who were given an ambient listening AI product (DAX CoPilot) after a 1 month baseline period and asked to use it for clinical documentation with a focus on problem-focused visits over a 3-month period, with a control group of 20 clinicians continuing traditional documentation methods. The primary outcomes include changes in documentation efficiency (measured through metrics such as time spent on documentation per patient) and clinician burnout (assessed using the validated Mini-Z 2.0 burnout inventory). Secondary outcomes involve patient satisfaction with clinicians' use of the AI tool and examining end-user technology acceptance among clinicians using a survey based on the Unified Theory of Acceptance and Use of Technology (UTAUT). The study aims to provide insights into the potential of AI-assisted documentation tools in enhancing clinical workflow and addressing the growing concern of clinician burnout.

Inclusion criteria

• Licensed Clinicians: Independently licensed clinicians (MDs, DOs, NPs, PAs) who have been actively practicing at Samaritan Health Services for at least 6 months.
• Primary Care Only: Providers must have a listed specialty of family medicine, internal medicine, or pediatrics, and currently practice primarily in a primary or urgent care clinic.
• Provider has an Apple iPhone and is willing to install Epic Haiku.

Exclusion criteria

• Inpatient-Only Clinicians: Exclude clinicians who only, or primarily, work in inpatient settings, as documentation needs and challenges may differ significantly from those in outpatient settings.
• Trainees: Exclude medical students and residents due to their varying levels of experience and dependence on supervisory oversight.
• Minimum Outpatient Encounters: Exclude clinicians with fewer than 100 outpatient encounters per month to focus on those with a significant workload in outpatient settings.
• Android Smartphones Users: Clinicians may not use Android, or generally any non-Apple or non-iOS smartphones, given the software limitations of the selected intervention technology.
• Corrective Action: Exclude clinicians facing dismissal, corrective, or disciplinary action.
• Scheduled leave longer than 3 weeks during the study period